vrr/koboldcpp
2
0
Fork 0
mirror of https://github.com/LostRuins/koboldcpp.git synced 2025-09-11 09:34:37 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

add ability to use guide tokens for TTS, ref: https://github.com/ggerganov/llama.cpp/pull/11186

Browse source
This commit is contained in:
Concedo 2025-01-11 17:18:21 +08:00
parent bd38665e1f
commit 07173e84a0
4 changed files with 53 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

4
src/llama.cpp
View file

@ -11122,9 +11122,9 @@ static int llama_decode_impl(
}
}
GGML_ASSERT(n_tokens_all <= cparams.n_batch);
GGML_ASSERT_CONTINUE(n_tokens_all <= cparams.n_batch);
GGML_ASSERT((cparams.causal_attn || cparams.n_ubatch >= n_tokens_all) && "non-causal attention requires n_ubatch >= n_tokens");
GGML_ASSERT_CONTINUE((cparams.causal_attn || cparams.n_ubatch >= n_tokens_all) && "non-causal attention requires n_ubatch >= n_tokens");
if (lctx.t_compute_start_us == 0) {
lctx.t_compute_start_us = ggml_time_us();